Opening up an API 550B...

[mutterd]

Hey guys,

any tips on getting the two circuit boards of an API 550B apart?

I've got one that's not passing signal…

thanks!
T.

 

[outoftune]

there is about a 95% chance the problem is with one of the two 2520 opamps. you don't need to take it apart to remove those. you can probe their inputs / outputs with a scope to see where the signal stops.

i have fixed dozens of these and this is the usual culprit. if both 2520 are fine, then it is likely the 2510 balanced input receiver.

 

[mutterd]

Yeah,

That was my thought as well...

My 550b is an older one and has 2 opamps, can only see one of them as the other is in between the PCB's...

I swapped the accessible one out with a known working one and there was no change, which is what brought me here, I couldn't find any easy way to split the two PCB's...

Was just wondering if there was a trick.

I'll check it out on the scope tonight to see what's up...

You don't happen to have a proper schematic for the 550b do you? All the legible ones I can find are for the 550a's.

Thanks for writing - you're here in Montreal, right?

T.

 

[joaquins]

We could assume the 550B is the same as 550A but with one more band, only adding one more voltage follower and switch, caps and resistors associated. Insert a signal as usual, then test the outputs of each opamp from input to output and see where the signal stops, could be with oscilloscope or a prove going to some monitoring, to listen the signal. Even if the opamp is between the two boards you may be able to get a prove to the output from the other side of the board. There are 3 opamps inside of it, 2 2520 and one 2510, which may fail also, and it's the input stage. At least this way you may find where the problem comes from and then find a way to fix it, but you already have it working. A friend of mine has a lot of API devices, never listen him talking about non working units, he always let one empty space over the lunchbox with a grid for better cooling, maybe if you don't is much easier to fail.

JS

 

[mutterd]

ok, so i've confirmed that the signal stops at the first opamp in the circuit - not sure if its a 2520 or a 2510 cause i can't see it at this point…

looks like i'll have to peel it open to get to it.

thanks guys.
T.

 

[joaquins]

IIRC the 2510 is completely different, physically, diff pinout and case, so it should be pretty easy to know. I've hear a lot of time of them blowing out, there is a company that made those on double boards, not potted, most components, if not all of them SMD, maybe they are a good choice, people say they are more reliable, less temperature problems, not original though. I don't remember who are they but someone would tell.

JS

 

[pucho812]

Your referring to the scott liebers 2510 replacement

http://scottliebers.com/op_amps/op_amps_2510.htm

 

[LRRec]

Are you sure the 550b has a balanced input?

The pair I had from the 90's were unbalanced like the 550a. No 2510, just 2520's.

 

[joaquins]

Are you sure the 550b has a balanced input?

The pair I had from the 90's were unbalanced like the 550a. No 2510, just 2520's.

No, I'm just saying if the first opamp is different than a 2520 it must be a 2510. I said that up there, when he said the first DOA was the problem I asked if it was like a 2520 or not. I've never seen a 550b inside, only 550a.

quick search, 550b is unbalanced input, 2520 should be replaced with newer ones (after 1990) since they are smaller and older won't fit between the boards, newer ones use SMD voltage followers.

"API 550b

4 band, 7 frequencies per band, 12 db's stepped boost/cut, discrete with LF and HF shelving on toggle switches
Proportional Q
Unbalanced input
uses two 2520 opamps

A well designed and versatile extension to the 550A, but a little different in sound to the 550A due to small changes in the Q. The 2520 opamps were on sockets, making it easy to replace on the two, double-sided circuit boards. The input 2520 can only be replaced with a new type (from 1990 on) because they are a slight bit smaller than the old ones, which do not physically fit due to an adjacent circuit board . Later 550b's used surface-mount VF amps."

The page also mention 550a1 with a mistake, it wasn't meant to be cheaper, IC opamps where an experiment, believing the IC opamps at that time were better than when 550 was released so they tried to, they found the sound wasn't as good as expected so they came back to discrete, all was socket and IIRC the PCB of those and 550a of that time where the same, they were designed to be any of those.

JS

 

[outoftune]

the 2510 is a totally different physical size. it is a single inline package and the official API ones are blue in color.

the 2520 follows the standard DOA pinout, and the ins/outs are clearly labelled on the opamp.

yes i am in montreal.

i have never seen a schematic for one, API doesn't release those. but it's either the 2510 or 2520's that are likely to cause the problem.

 

[mutterd]

So my model only had two 2520's… were there models of the 550b that had the 2520's and the 2510's?

the 2 boards are connected by a header - if you remove the knobs and knurled nuts the two circuit boards just come apart.

the 2520 that is in-between the two boards has shorter legs so its lower profile. I didn't have the heart to trim down any of my working 2520's, but a Gar2520 from classic api fit perfect!

so my first session with it is tonight - so far so good!

thanks for all the help guys.
T.

 

[rubli]

hello, I have a 550B with a non working band (low mid) I took it apart to find something obvious, but I did not see anything starnge, maybe a defective switch or an open path ?anyone can shed some light on this problem ? I could not locate a schematic, would be very handy... thank folks

 

[Whoops]

hello, I have a 550B with a non working band (low mid) I took it apart to find something obvious, but I did not see anything starnge, maybe a defective switch or an open path ?anyone can shed some light on this problem ? I could not locate a schematic, would be very handy... thank folks

The schematic can be found in the Technical Docs section of this forum:
https://groupdiy.com/threads/api.44738/page-2
If it's just one band that is not working then located that band in the schematic and focus just on that part of the circuit

 

[pucho812]

hello, I have a 550B with a non working band (low mid) I took it apart to find something obvious, but I did not see anything starnge, maybe a defective switch or an open path ?anyone can shed some light on this problem ? I could not locate a schematic, would be very handy... thank folks

Possibly a faulty capacitor

 

[Ike Zimbel]

Possibly a faulty capacitor

Yes, very possible. I think every "vintage" 550 (A,B, the mysterious old "550M", not the recent "Mastering" version) that I have ever been into has had electrolytic caps in very bad shape. The 16v decoupling caps on the 16v supply rails are virtually always toast. Replace those with a higher voltage cap, 25v> (and make note of the orientation before your remove them! I have a vague recollection of the "+" marking on either the parts overlay or the actual circuit board being incorrect in one position).

 

[FIX]

The page also mention 550a1 with a mistake, it wasn't meant to be cheaper, IC opamps where an experiment, believing the IC opamps at that time were better than when 550 was released so they tried to, they found the sound wasn't as good as expected so they came back to discrete, all was socket and IIRC the PCB of those and 550a of that time where the same, they were designed to be any of those.

JS

THe 550A-1 wasn't a mistake, it was an attempt from API, right before they went out of business to make the A cheaper. At the time everyone had forgotten the tech behind a passive filter, where you had to keep the over all impedance constant if you wanted the frequency to stay the same. That is why just about every EQ in those days was a Proportional Q circuit, the constant impedance altered the Q. When I started at API (Datatronix) they inherited the -1, not realizing that it wasn't proportional Q, and started selling them. When Sunset Sound sounded the alarm to us, I pulled out the old HP spectrum analyzer and quickly realized that they were fixed Q. All of us, including Saul dropped our jaws. That brought about the re-design of the A with the connectorized switches, just so they could be tested. We also decided to make the buffer card the same pinout as the TL084 op-amp so you could swap one out for testing or in an emergency.

The 550A-1 is basically a 560A graphic with rotary switches. It isn't a bad EQ, it's just not what you expect it to be.

 

[FIX]

Possibly a faulty capacitor

Hardly ever a faulty cap, it would still work to some degree, most likely the Sansie 6 pole switch. They are getting to that age. Could also be a buffer, but that would tale out 2 bands. You can fine schematics if you search, slight possibility that it could be one of the resistor ladders on the ceramic. If you have a microscope camera, look at each terminal and see if one has come loose. They can be desoldered.

 

[tomas.borgstrom]

I have replaced a faulty cap in my 550b. It's the remake, about 10-15 years old.